H. S. Teoh:
Combinatorial puzzles come to mind (Rubik's cube solvers and
its ilk,
for example). Maybe other combinatorial problems that require
some kind
of exhaustive state space search. Those things easily go past
20! once
you get beyond the most trivial cases.
I know many situations/problems where you have more than 20!
cases, but in those problems you don't iterate all permutations,
because the program takes ages to do it. In those programs you
don't use nextPermutation, you scan the search space in a
different and smarter way.
I don't know of any use case for permuting so large sets of items.
Bye,
bearophile
